Activities Related to Credit Intermediation (NAICS 2022 Code 5223)
A Histometrics industry-group primer for public- and private-market investors
1. Overview
The North American Industry Classification System (NAICS) code 5223 is the "fees-and-plumbing" tier of the credit system. It sits inside subsector 522 (Credit Intermediation and Related Activities) alongside the banks that take deposits and make loans (industry group 5221) and the finance companies that lend without deposits (5222). What sets 5223 apart is what its firms do not do: they do not take deposits, and they mostly do not lend their own money or carry loans on a balance sheet. Instead they earn fees and spreads for services around credit and payments — arranging loans, moving and settling money, and collecting payments on loans other people own.[1]
That single distinction drives the economics of the whole group. Because these firms earn fees rather than interest, they carry far less credit risk than a bank or a lender, need less funding capital, and — at the payments end — enjoy some of the most attractive network economics in all of finance. The trade-offs are heavy regulation, intense fee competition, and, in places, sharp cyclicality. But 5223 is not one business: it bundles together three very different industries whose sizes, growth paths, owners, and margins diverge widely. The value of reading it at this level is the contrast among the three — so we lead with that.
2. What's inside — and how the three children differ
NAICS narrows from broad to specific: sector (2-digit) → subsector (3-digit) → industry group (4-digit) → industry (5-digit) → national industry (6-digit). Code 5223 is a four-digit industry group with three five-digit children. Each is a "single-child" line (its five- and six-digit codes are identical), so the names below double as the six-digit activities.
| Child (5-digit) | What it does | ~Share of group receipts | Direction of travel | Who owns it | How you invest |
|---|---|---|---|---|---|
| 52232 Financial Transactions Processing, Reserve & Clearinghouse | Moves, clears and settles payments — card networks, processors/acquirers, ACH & ATM operators, clearinghouses | ~76% | Structural grower (cash-to-digital) | Large public companies + member-owned bank utilities + private fintech | Deep public market; some infra private/member-only |
| 52239 Other Activities Related to Credit Intermediation | Loan servicing (mortgage, student) + money transmission/remittances + check cashing, money orders | ~15% | Mixed — servicing steady, remittances grow but compress, cash access shrinks | Nonbank public servicers + public/private money-transfer + small cash-access shops | Public by sub-segment; some private |
| 52231 Mortgage & Nonmortgage Loan Brokers | Matches borrowers to lenders for a commission; funds nothing, holds nothing | ~10% | Cyclical, structurally gaining share from retail lenders | Overwhelmingly small, private, owner-operated | Almost no pure public play; mostly private |
Three things jump out. First, money movement (52232) is the whole story on dollars — roughly three-quarters of the group's receipts — while the two credit-adjacent industries split the rest. Second, the direction of travel differs: payments is a durable secular grower riding the shift from cash to digital; servicing/money-transfer is a mixed bag; brokering is a rate-driven cyclical. Third, the ownership models are almost opposites: payments concentrates in large listed firms (plus a few systemically important private utilities), whereas brokering is a long tail of tiny private shops with barely any public footprint. An investor "buying 5223" is really choosing among a network toll-taker, a fee-services book, and a leveraged bet on loan volume — very different animals under one code.[4][5][6]
The group deliberately excludes the activities that look adjacent but sit elsewhere: the deposit-taking banks (5221), the lenders that actually fund mortgages and consumer loans (5222, e.g. Real Estate Credit 522292, where names like Rocket and UWM's lending arm are classified), the Federal Reserve's own clearing (521110), and securities-market clearinghouses such as The Depository Trust & Clearing Corporation, DTCC (adjacent investment codes). The test throughout is who takes the credit risk and holds the balance — if a firm does, it is a bank or a lender and lives in 5221/5222, not here.[1][4][5]
3. How big it is (this group's rollup figures)
Federal figures for NAICS 5223, from our ground-truth file for this level: receipts, firm counts and concentration are 2022 Economic Census; establishments, employment and payroll are 2023 County Business Patterns (CBP). The three children's figures aggregate almost exactly to the group totals — a useful cross-check.[2][3]
| Metric | NAICS 5223 (group) | Source (year) |
|---|---|---|
| Receipts (revenue) | ~$179.9 billion | Economic Census (2022)[2] |
| Firms | 16,131 | Economic Census (2022)[2] |
| Establishments | 25,803 | County Business Patterns (2023)[3] |
| Paid employees | 310,591 | County Business Patterns (2023)[3] |
| Annual payroll | ~$35.28 billion | County Business Patterns (2023)[3] |
| First-quarter payroll | ~$10.0 billion | County Business Patterns (2023)[3] |
Concentration (group). The four largest firms took 35.3% of receipts (the four-firm concentration ratio, CR4), the top eight 49.4% (CR8), the top 20 61.2% (CR20) and the top 50 73.6% (CR50); the Herfindahl-Hirschman Index (HHI, which sums every firm's squared market share and runs from near 0 for a highly fragmented market to 10,000 for a monopoly) was 388.1 — technically "unconcentrated" under U.S. antitrust thresholds (below 1,000).[2][17]
But the group HHI hides the real structure. Because 5223 mixes three unlike industries, its blended HHI (388) sits between a concentrated payments industry and two fragmented ones — and understates concentration inside each activity. Where the money is (payments, 52232) is the most concentrated child (HHI ~660, CR4 46.5%); the two smaller children are far more fragmented (loan brokers HHI ~218, CR4 24.3%; other credit activities HHI ~263, CR4 25.4%).[4][5][6] And even those child HHIs understate reality, because money transfer and mortgage servicing are each led by a handful of names within their own sub-segments.
The children also differ starkly in firm scale, which the receipts-per-establishment column makes vivid:
| Child | Receipts | Establishments | Employees | ~Receipts per establishment | ~Average pay | HHI |
|---|---|---|---|---|---|---|
| 52232 Payments/clearing | ~$136.7B | 4,711 | 177,887 | ~$29.0M | ~$133K | ~660 |
| 52239 Other credit activities | ~$26.1B | 11,300 | 77,752 | ~$2.3M | ~$87K | ~263 |
| 52231 Loan brokers | ~$17.1B | 9,792 | 54,952 | ~$1.75M | ~$89K | ~218 |
Payments establishments are, on average, roughly 16 times larger by revenue than a broker's storefront and pay the highest wages in the group — a few thousand high-productivity firms versus tens of thousands of small ones. That single contrast explains why 18% of the group's establishments (52232's) generate 76% of its receipts.[4][5][6]
Undercount caveat — large, and concentrated where small/individual ownership dominates. The $179.9 billion counts only the receipts of employer firms with paid (W-2) staff, and it understates the true footprint in three different ways, one per child:
- Loan brokers (52231): the mortgage business runs on self-employed, commission-only (1099) originators and nonemployer sole proprietors that CBP and the Economic Census omit. Federal counts show ~55,000 paid employees here, but licensing data recorded roughly 221,000 active mortgage loan originators nationally in 2025 — the surveys catch a fraction of the people actually doing the work.[4]
- Other credit activities (52239): the largest servicing books and most money-transfer revenue sit outside this code. PennyMac services $716.6 billion of mortgages and Nelnet services $434.5 billion of federal student loans, but those revenues are reported by firms classified elsewhere; Western Union alone booked about $4.2 billion in 2024, most of it earned abroad; and tiny cash-access shops are largely missed.[6][11][12][10]
- Payments/clearing (52232): the $136.7 billion is the fee pool of private payment vendors, not the value of U.S. payments. It excludes the Federal Reserve's clearing (FedACH, Fedwire, FedNow, in 521110), the securities clearinghouses (DTCC alone processed on the order of $3.7 quadrillion in 2024), and banks' in-house card processing. The dollars actually cleared and settled run into the quadrillions annually.[5]
Read $179.9 billion as "the standalone employer fee pool of these three activities," not the economic weight of the credit-support and payments machine, which is far larger.
4. Investable universe — where value concentrates across the children
Value is heavily lopsided toward payments (52232) — the child that holds three-quarters of receipts also holds the great majority of the group's listed market value, and the highest-quality economics.
- Payments / clearing (52232) — the center of gravity. The two global card networks, Visa (V) and Mastercard (MA), are capital-light transaction switches with operating margins above 60%. Around them sit scaled processors and merchant acquirers — Fiserv (FISV), FIS (FIS), Global Payments (GPN, which absorbed Worldpay in January 2026) — fintech platforms PayPal (PYPL), Block (XYZ), Shift4 (FOUR), Marqeta (MQ); money-transfer/EFT names Western Union (WU), Euronet (EEFT), Corpay (CPAY); and niche processors Jack Henry (JKHY), WEX (WEX), ACI Worldwide (ACIW), EVERTEC (EVTC). Crucially, much of the most systemically important infrastructure is not listed: member-owned bank utilities (The Clearing House, which runs CHIPS, the private ACH network and the RTP real-time rail; Early Warning Services, operator of Zelle), plus private fintech Stripe.[5]
- Other credit activities (52239) — two scaled sub-baskets. In loan servicing: nonbank mortgage servicers Rocket/Mr. Cooper (RKT), PennyMac (PFSI), Rithm/Newrez (RITM), Onity (ONIT), and student-loan servicer Nelnet (NNI). In money transmission: Western Union (WU), Euronet (EEFT), Wise (WISE), Remitly (RELY), Intermex (IMXI). No single listed name is a clean read-through to the federal statistics; investors use a basket of imperfect proxies.[6]
- Loan brokers (52231) — almost no pure public play. True brokers are small and private, so listed exposure is indirect: a franchise platform (RE/MAX Holdings, RMAX, via its Motto Mortgage brokerage franchise), online loan/lead marketplaces (LendingTree, TREE; NerdWallet, NRDS), and the wholesale lender that sells only through brokers (UWM Holdings, UWMC — a broker-channel proxy, not itself a broker). The real ownership opportunity is private.[4]
Prices, yields and multiples vary widely across these groups and should be checked at the diligence date.
5. How the money works
Every owner in 5223 monetizes flows and services, not balances — so bank-style net interest margin does not apply. But the fee model differs by child:
- Loan brokers (52231): a fee per closed loan and no interest spread. On mortgages the commission is typically ~1%–2.75% of the loan amount (lender-paid or borrower-paid, never both, never tied to the rate). Revenue is funded volume against a largely fixed cost base — high operating leverage, very profitable in booms, loss-making when volume collapses, with almost no balance-sheet risk.[4]
- Payments/clearing (52232): a slice of payment volume and a fee per transaction. Card networks earn service, data-processing and cross-border fees on volume (they do not keep interchange, which goes to issuing banks); acquirers keep a net take rate in basis points after passing interchange through; wallets and money transmitters add float income on balances (rate-sensitive) and an FX spread; clearing utilities charge per-transaction and membership fees near cost. The through-line is network effects and operating leverage — the rails are built, so incremental volume drops to profit at very high rates.[5]
- Other credit activities (52239): servicers earn a servicing fee — roughly 0.25% a year (25 basis points) of the loan's unpaid principal balance (UPB) — plus ancillary fees and escrow "float." Their core asset is the Mortgage Servicing Right (MSR), whose value rises when rates rise (slower prepayments) and falls when rates drop. Money-transfer providers earn a transaction fee plus an FX spread on send volume × a take rate; check cashers and money-order issuers earn per-item fees shaped by state caps.[6]
The unifying idea: these are fee and agency businesses with low credit risk, which is exactly what separates 5223 from the deposit-taking and lending industries above it in subsector 522. The exception is the servicer, whose MSR is a genuine balance-sheet asset with real interest-rate sensitivity.
6. Demand drivers
Because the three children respond to different forces — sometimes in opposite directions — the group has no single demand switch:
- Interest rates move brokering and servicing in opposite directions: lower rates trigger refinancing that lifts broker volume but shortens servicing cash flows and cuts MSR values; higher rates freeze originations but make servicing books more valuable.[4][6]
- Nominal spending and the cash-to-digital shift drive payments: most fees are percentage-based, so inflation lifts revenue even when real volume is flat, and the multi-decade migration from cash keeps compounding transactions (the Federal Reserve's 2025 study counted 236.6 billion U.S. noncash payments worth $140.0 trillion in 2024, with the longest runway abroad).[5][9]
- The stock of debt outstanding (not new loans) drives servicing revenue, making it far steadier than origination.[6]
- Migration, employment and immigrant wages drive remittances (biggest U.S. corridors: Mexico, India, Central America, the Philippines).[6]
- Structural share shifts: independent brokers keep taking share from retail lenders (broker channel now ~20–25% of originations), and digital adoption grows the money-transfer pie while squeezing per-transfer economics.[4][6]
7. Regulation
Oversight across 5223 is activity-based, not tied to the NAICS code, and heavily dual federal-and-state. The regimes differ by child:
- Loan brokers: the Consumer Financial Protection Bureau (CFPB) enforces the SAFE Act (mortgage-originator licensing through the Nationwide Multistate Licensing System, NMLS), the Loan Originator Compensation Rule (Regulation Z under the Truth in Lending Act, TILA — pay can't depend on loan terms), and RESPA (Regulation X, the Real Estate Settlement Procedures Act — no kickbacks), on top of 50 state licensing regimes.[4]
- Servicing / money transfer: mortgage servicers fall under RESPA/TILA, CFPB servicing rules and government-sponsored-enterprise (Fannie Mae, Freddie Mac, Ginnie Mae) eligibility. Money transmitters must register with the Financial Crimes Enforcement Network (FinCEN) as a Money Services Business, run a Bank Secrecy Act / anti-money-laundering (BSA/AML) program, and hold state Money Transmitter Licenses. A 2026 change: the One Big Beautiful Bill Act imposed a 1% federal excise tax on cash-funded remittances from January 1, 2026, hitting the agent/cash model hardest.[6][16]
- Payments: the Federal Reserve's Regulation II (Durbin Amendment) on debit interchange (its fee cap vacated in the 2025 Corner Post ruling, now on appeal); CFPB Regulation E; SEC/CFTC clearing-agency oversight and Title VIII supervision of systemically important utilities; private card-network rules and PCI DSS data-security standards; live antitrust (the DOJ's 2024 Visa suit); and the new GENIUS Act (signed July 2025) creating a federal framework for payment stablecoins.[5][19]
8. Consolidation
The three children are consolidating on different logics:
- Payments (52232) consolidates toward scale and network effects — the biggest wave of the decade: Fiserv–First Data, FIS–Worldpay and Global Payments–TSYS (2019), the three-way Global Payments / Worldpay / FIS reshuffle that closed January 2026, and Capital One's ~$35 billion acquisition of Discover (May 2025), which created a third vertically integrated network-plus-issuer and nudged the Visa/Mastercard duopoly toward a "triopoly."[5]
- Servicing/money transfer (52239) consolidates toward scaled books: Rocket's $14.2 billion acquisition of Mr. Cooper (closed October 1, 2025) combined the largest originator and servicer into a book covering nearly 10 million homeowners; PennyMac agreed to buy Cenlar's subservicing; in money transfer, MoneyGram went private (2023) and Western Union agreed to acquire Intermex (pending in 2026) even as digital-first Wise and Remitly pressure incumbents' take rates.[6][13]
- Loan brokers (52231) stay fragmented — the field remains crowded (HHI ~218), with the only notable trend being the rise of "mega-brokers" recruiting originators for local scale.[4]
The net effect is why the group's CR4 sits at only 35.3% even though its largest activity is meaningfully concentrated: two of the three children are structurally fragmented.
9. Risks
- Interest-rate whiplash cuts both ways within the group: a sharp rate drop lifts broker volume but forces MSR write-downs at servicers; a spike does the reverse.[4][6]
- Rate and volume cyclicality at brokers, whose thin margins and high operating leverage turn small volume swings into large profit swings.[4]
- Fee / take-rate compression — debit interchange (Durbin, Corner Post), merchant litigation, and digital money-transfer players all squeeze fees.[5][6]
- Disintermediation — real-time account-to-account (A2A) rails, regulated stablecoins and AI-driven ("agentic") commerce could route volume around card incumbents over time.[5]
- Regulatory, political and litigation risk — servicing rules, student-loan policy, the new remittance excise tax, the DOJ Visa suit, and originator-compensation/RESPA enforcement all shift the economics.[5][6][16]
- Operational, cyber, fraud, AML and settlement risk — these firms are critical infrastructure; a clearinghouse outage is a systemic event, and BSA/AML or sanctions failures carry large fines.[5][6]
- Data undercount — employer-only federal statistics materially understate the true number of small brokers and cash-access operators.[3]
10. How to invest & outlook
Invest by child, not by the four-digit code — the three activities behave differently enough that a single "5223" thesis would blur them.
- Public routes concentrate in payments (52232), spanning the risk spectrum: network compounders with toll-taker economics and premium multiples (Visa, Mastercard); cheaper, more cyclical scaled processors (Fiserv, FIS, Global Payments); higher-volatility fintech platforms (PayPal, Block, Shift4, Marqeta); and value/yield money transfer (Western Union, Euronet) — plus thematic payments/fintech exchange-traded funds (ETFs) for diversified exposure. Servicing (52239) is a rate/MSR bet (Rocket, PennyMac, Rithm, Onity, Nelnet). Brokering (52231) has no pure public play; use indirect proxies (RMAX via Motto, TREE, UWMC).
- Private routes reach what the public market can't: the member-owned utilities (DTCC, The Clearing House — effectively member-only), late-stage/secondary fintech (Stripe the marquee name), and — at the small end — local brokerages, mortgage subservicing, and lending against servicing rights and receivables.
One discipline applies everywhere: never equate a firm's payment or loan volume with its revenue or with Census receipts. A network's gross payment volume, a lender's origination figure and a servicer's UPB are all far larger than the fee income this code actually measures.
Outlook. The three trajectories from Section 2 still hold. Payments is the durable structural grower — picks and shovels to the whole economy's spending — with regulation (the Corner Post/Durbin fight, GENIUS Act stablecoins) and technological disruption (A2A, agentic commerce) the two variables that decide how much of the toll incumbents keep. Servicing offers steadier, stock-of-debt revenue whose swing factors are the rate path and digestion of the Rocket–Mr. Cooper combination; money transfer grows in volume but compresses in take rate and now carries a new remittance tax. Brokering is a cyclical recovery play — forecasters see the 30-year mortgage rate near 6.4%–6.5% through 2026 and a gradual, purchase-driven rise in originations — layered on a durable structural tailwind as brokers keep taking share from retail. The need for credit intermediation is permanent; how each of these three fee businesses earns from it is what an investor is really choosing among.
Sources
- U.S. Census Bureau / Office of Management and Budget. "2022 NAICS — Industry Group 5223 (Activities Related to Credit Intermediation) and children 52231, 52232, 52239: definitions, scope and cross-references." 2022. https://www.census.gov/naics/?input=5223&year=2022
- U.S. Census Bureau. "2022 Economic Census — NAICS 5223: receipts, firms, concentration ratios (CR4/CR8/CR20/CR50) and HHI" (Histometrics ground-truth compilation for this level). 2022. https://www.census.gov/programs-surveys/economic-census.html
- U.S. Census Bureau. "2023 County Business Patterns — NAICS 5223: establishments, employment, annual and first-quarter payroll." 2023. https://www.census.gov/programs-surveys/cbp.html
- Histometrics. Industry primer — NAICS 52231, Mortgage and Nonmortgage Loan Brokers (definition, size, broker economics, ~221k licensed originators, investable proxies, MBA and HousingWire commentary). 2026.
- Histometrics. Industry primer — NAICS 52232, Financial Transactions Processing, Reserve, and Clearinghouse Activities (networks/processors/utilities, receipts $136.7B, concentration, regulation, consolidation). 2026.
- Histometrics. Industry primer — NAICS 52239, Other Activities Related to Credit Intermediation (loan servicing and money transmission, receipts $26.1B, MSR economics, remittance tax). 2026.
- Mortgage Bankers Association. "Mortgage Finance Forecast — 2026 origination volumes (~$2.2 trillion) and channel commentary." Oct. 2025. https://www.mba.org/news-and-research/newsroom
- HousingWire. "With ~20% market share, independent mortgage brokers are competing directly with retail lenders." 2024. https://www.housingwire.com/articles/with-20-market-share-independent-mortgage-brokers-are-competing-directly-with-retail-lenders/
- Federal Reserve Board. "2025 Triennial Payments Study — Initial Findings (236.6 billion noncash payments worth $140.0 trillion in 2024)." 2026. https://www.federalreserve.gov/newsevents/pressreleases/other20260701a.htm
- The Western Union Company. "Form 10-K, FY2024 (revenue ~$4.2B; ~600,000 agent locations)." 2025. https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001365135&type=10-K
- PennyMac Financial Services, Inc. "Form 10-K, FY2025 (servicing portfolio $716.6B UPB)." 2026. https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001745916&type=10-K
- Nelnet, Inc. "Form 10-K, FY2025 ($434.5B federal student loans serviced for 11.4M borrowers)." 2026. https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001258602&type=10-K
- Rocket Companies, Inc. "Rocket Companies Closes $14.2 Billion Acquisition of Mr. Cooper." 2025. https://www.rocketcompanies.com/press-release/rocket-companies-closes-14-2-billion-acquisition-of-mr-cooper/
- Capital One Financial Corp. "Capital One Completes Acquisition of Discover Financial Services (~$35B, completed May 2025)." 2025. https://investor.capitalone.com/news-releases/
- U.S. Congress / U.S. Treasury. "GENIUS Act — federal framework for payment stablecoins (signed July 18, 2025)." 2025. https://www.congress.gov/
- Internal Revenue Service. "Proposed regulations on the new remittance-transfer excise tax under the One Big Beautiful Bill Act (1%, effective Jan. 1, 2026)." 2025. https://www.irs.gov/newsroom/treasury-irs-issue-proposed-regulations-on-the-new-remittance-transfer-tax-established-under-the-one-big-beautiful-bill
- U.S. Department of Justice / Federal Trade Commission. "Herfindahl-Hirschman Index and Merger Guidelines concentration thresholds (unconcentrated below 1,000)." 2023. https://www.justice.gov/atr/herfindahl-hirschman-index
- Consumer Financial Protection Bureau; Financial Crimes Enforcement Network. "Mortgage servicing rules (RESPA/Reg X, TILA/Reg Z); SAFE Act / NMLS originator licensing; Money Services Business registration and BSA/AML." 2023–2026. https://www.consumerfinance.gov/; https://www.fincen.gov/
- U.S. Department of Justice. "United States v. Visa Inc. — civil antitrust complaint (debit-network monopolization)." 2024. https://www.justice.gov/opa/pr/justice-department-sues-visa-monopolizing-debit-markets